	**********************************************************************
	* File Name:	secondary_trauma_ps_replication.do		             *
	* Date:   		June 5, 2025			    		                 *
	* Authors: 		Marty P. Jordan, Elinor R. Jordan, & Lauren S. Foley *
	*                                                                    *
	*                                                                    *
	* Contact:      Marty P. Jordan (jordan61@msu.edu)                   *
	*                                                                    *
	* Purpose:		This Do file creates the analysis evaluating         *
	*               secondary trauma experienced by political science    *
	*               undergraduate and graduate student interns at        *
	*               WMU and MSU.                                         *
	*                                                                    *
	* Input Files:	secondary_trauma.xlsx                                *				
	* 				                    								 *
	* Output File:	secondary_trauma_ps_replication.log,               	 *				
	*               secondary_trauma_ps_replication.dta                  *			
	*  				                                                     *
	* Version:		Stata/SE 14.2					                     *
	**********************************************************************	

	**********************************************************************
	* Start log file													 *
	**********************************************************************

log using secondary_trauma_ps_replication.log, replace text	
	
	*************************
	* Specs and Preferences *
	*************************
	
version 14.2
set more off
quietly log
set seed 07312004

	******************************
	* Install necessary packages *
	******************************

ssc install extremes, replace
ssc install estout, replace
ssc install outreg2, replace
ssc install bacondecomp, replace
ssc install coefplot, replace
ssc install grc1leg2, replace
net install grc1leg, from("http://www.stata.com/users/vwiggins") replace
net install sg113_2, from ("http://www.stata-journal.com/software/sj9-4") replace
net install st0039, from ("http://www.stata-journal.com/software/sj3-2") replace
net install st0004, from ("http://www.stata-journal.com/software/sj1-1") replace
net install sg164_1, from ("http://www.stata-journal.com/software/sj8-1") replace
net install st0094, from ("http://www.stata-journal.com/software/sj5-4") replace

	***********************************************************
	* Import respondent survey data                           *
	*								                          *
	* IRB Approval: STUDY00009990 & IRB-2023-354  			  *
	*                                                         * 
	* Any variables that might violate respondent privacy     *
	* or lead to identifying respondent have been removed     *
	* from the dataset that is made publicly available.       *
	***********************************************************

import excel using secondary_trauma_ps_replication, sheet("vars") firstrow case(lower)	
	
	********************************************
	* Label variables and provide descriptions *
	********************************************
	
label variable	post	"post_survey=1, pre_survey=0"
label variable	finished	"finished survey=1, 0=otherwise"
label variable  time "number of times survey completed"
label variable	id	"id"
label variable	msu	"attends msu=1, attends wmu=0"
label variable	intern_type	"internship placement type, 1=advocacy/lobbying; 2=judicial/legal; 3=executive; 4=legislative/Congress; 5=local; 6=nonprofit; 7=campaign/political party; 8=other"
label variable	paid	"paid internship=1, unpaid=0"
label variable	hours	"internship hours, on average, worked per week"
label variable	intern_prior	"previously done internship=1, otherwise=0"
label variable	intern_prior_number	"number of prior internships"
label variable	primary_trauma	"experienced primary trauma, Extensive exposure=5, no exposure=1"
label variable	secondary_trauma_heard	"previously heard term secondary trauma=1, maybe=0.5, no=0"
label variable	secondary_trauma_exp	"experienced secondary trauma=1, maybe=0.5, no=0"
label variable	secondary_trauma_anticipate	"degree of secondary trauma anticipated in internship, 5=definitely to 1=definitely not"
label variable	symptoms_numb	"STS symptoms (5=very often, 1=never): I felt emotionally numb, "
label variable	symptoms_heart	"STS symptoms (5=very often, 1=never): My heart started pounding when I thought about my work with clients, constituents, or customers"
label variable	symptoms_reliving	"STS symptoms (5=very often, 1=never): It seemed as if I was reliving the trauma(s) experienced by my clients, constituents, or customers"
label variable	symptoms_sleep	"STS symptoms (5=very often, 1=never): I had trouble sleeping"
label variable	symptoms_despondent	"STS symptoms (5=very often, 1=never): I felt discouraged about the future"
label variable	symptoms_upset	"STS symptoms (5=very often, 1=never): Reminders of my work with clients, constituents, or customers upset me"
label variable	symptoms_antisocial	"STS symptoms (5=very often, 1=never): I had little interest in being around others"
label variable	symptoms_jumpy	"STS symptoms (5=very often, 1=never): I felt jumpy"
label variable	symptoms_lessactive	"STS symptoms (5=very often, 1=never): I was less active than usual"
label variable	symptoms_preoccupied	"STS symptoms (5=very often, 1=never): I thought about my work with clients, constituents, or customers when I didn't intend to"
label variable	symptoms_distracted	"STS symptoms (5=very often, 1=never): I had trouble concentrating"
label variable	symptoms_avoidance	"STS symptoms (5=very often, 1=never): I avoided people, places, or things that reminded me of my work with clients, constituents, or customers"
label variable	symptoms_dreams	"STS symptoms (5=very often, 1=never): I had disturbing dreams about my work with clients, constituents, or customers"
label variable	symptoms_avoidsome	"STS symptoms (5=very often, 1=never): I wanted to avoid working with some clients, constituents, or customers"
label variable	symptoms_annoyed	"STS symptoms (5=very often, 1=never): I was easily annoyed"
label variable	symptoms_fear	"STS symptoms (5=very often, 1=never): I expected something bad to happen to me"
label variable	symptoms_memory	"STS symptoms (5=very often, 1=never): I noticed gaps in my memory about client, constituent, or customer sessions"
label variable	physical_eatregularly	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Eat regularly (e.g., breakfast and lunch)"
label variable	physical_eathealthy	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Eat healthfully"
label variable	physical_exercise	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Exercise or go to the gym"
label variable	physical_healthcheckup	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Get regular medical care for prevention"
label variable	physical_sicktime	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Take time off when you are sick"
label variable	physical_sleep	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Get enough sleep"
label variable	physical_notechnology	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Get away from stressful technology, such as phones, tablets, and computers"
label variable	physical_other	"How often do you engage in the following PHYSICAL self-care (5=very often, 1=never)?: Other:"
label variable	physical_other_text	"How often do you engage in the following PHYSICAL self-care?: Other: - Text"
label variable	psychological_therapy	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Go to see a therapist or counselor for yourself"
label variable	psychological_stress	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Take a step to decrease stress in your life"
label variable	psychological_dreams	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Notice your inner experience---your dreams, thoughts, imagery, and feelings"
label variable	psychological_cultural	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Engage your intelligence in a new area---go to an art museum, performance, sports event, exhibit, or other cultural event"
label variable	psychological_sayno	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Say no to extra responsibilities"
label variable	psychological_outdoors	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Spend time outdoors"
label variable	psychological_other	"How often do you engage in the following PSYCHOLOGICAL self-care (5=very often, 1=never)? Other:"
label variable	psychological_other_text	"How often do you engage in the following PSYCHOLOGICAL self-care? - Other: - Text"
label variable	emotional_connect	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Stay in contact with important people in your life"
label variable	emotional_outrage	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Express your outrage in a constructive way"
label variable	emotional_selftalk	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Treat yourself kindly (supportive inner dialogue or self-talk)"
label variable	emotional_comforts	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Identify and seek out comforting activities, objects, people, relationships, places"
label variable	emotional_cry	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Allow yourself to cry"
label variable	emotional_laugh	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Find things that make you laugh"
label variable	emotional_emotions	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Process emotions by opening up to others"
label variable	emotional_other	"How often do you engage in the following EMOTIONAL self-care  (5=very often, 1=never)? Other:"
label variable	emotional_other_text	"How often do you engage in the following EMOTIONAL self-care? - Other: - Text"
label variable	professional_chat	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)?  Take time to chat with coworkers"
label variable	professional_time	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)? Make time to complete tasks"
label variable	professional_growth	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)? Identify projects or tasks that are exciting, growth promoting, and rewarding for you"
label variable	professional_support	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)? Get support from colleagues who have similar experiences"
label variable	professional_peer	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)? Have a peer support group"
label variable	professional_other	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care  (5=very often, 1=never)? Other:"
label variable	professional_other_text	"How often do you engage in the following WORKPLACE or PROFESSIONAL self-care? - Other Text"
label variable	class	"What is your class standing? 1=firstyear, 5=graduate student"
label variable	age	"What is your current age (in years)"
label variable	status	"What is your enrollment status at your university? 1=instate, 2=out-of-state, 3=international"
label variable	gpa	"What is your current cumulative GPA?"
label variable	minority	"racial/ethnic minority=1, 0=otherwise"
label variable	female	"identify as female=1, otherwise=0"
label variable	firstgen	"first-generation college student=1, otherwise=0"
label variable	financialaid	"Receive need-based financial aid=1, otherwise=0"
label variable	satisfaction	"How satisfied were you with your internship placement? Extremely satisfied=5, Extremely dissatisfied=1"
label variable	valuable	"Evaluate your internship experience this semester? Extremely valuable both professionally and personally=4, Not fulfilling for both my career and personal growth=1"
label variable	secondary_trauma_familiar	"How familiar are you with the term and concept of secondary trauma? Extremely familiar=5, not at all familiar=1"
label variable	actionplan	"Did you create a secondary trauma action plan? Yes=1, no=0"


	******************
	* BEGIN ANALYSES *
	******************

	*************************
	*** TABLE A1 APPENDIX ***
	*************************

	***********
	*** Explore descriptive statistics for the student and internship
	*** characteristics. This is for pre and post together. 
	***********

sum paid hours intern_prior intern_prior_number primary_trauma ///
class age status gpa minority female firstgen financialaid

	***
	*** RESULTS: Among survey respondents (n=146), 44% held paid internships, working 
	*** an average of 19 hours per week. Most respondents were juniors or seniors, 
	*** with an average age of 21 and a mean GPA of 3.67. Additionally, nearly 
	*** 40% identified as racial/ethnic minorities, over 60% as female, more 
	*** than 25% as first-generation college students, and over half received 
	*** financial aid.
	*** 

	***
	*** Produce the table for Appendix Table A1 containing descriptive statistics
	*** of respondents. This is for the pre survey descriptive stats for 
	*** Respondents
	***
	
estpost summarize class age status gpa female minority firstgen financialaid ///
	paid hours intern_prior intern_prior_number primary_trauma if post==0
	
esttab using tableA1.rtf, replace varwidth(38)	///
	cells("mean(fmt(2)) sd(fmt(2)) min(fmt(2)) max(fmt(2)) count(fmt(0))")  ///
	title("{\b Table A1: Descriptive Statistics of Survey Respondents}") ///
	varlabels (class "R's Class: Freshman (1) to Master's (6)" age ///
	"R's Age" status "R's In-State(1), Out-of-St.(2), Intl.(3)" ///
	gpa "R's GPA" female "Female R (1), Other (0)" minority ///
	"R Racial/Ethnic Minority (1), No (0)" firstgen ///
	"R First Gen. College Student (1), No (0)" financialaid ///
	"R's Fed. Financial Aid (1), No (0)" ///
	paid "R Paid Internship (1), No (0)" ///
	hours "R's Avg. No. Intern Hours/Week" ///
	intern_prior "R Completed Prior Internship (1), No (0)" ///
	intern_prior_number "R's No. of Prior Internships if Prior" ///
	primary_trauma "R's Primary Trauma, Extensive (5) None (1)") ///
	collabels("Mean" "SD" "Min" "Max" "N") ///
	compress nomtitle nonote noobs nonumbers label eqlabels(none) ///
	modelwidth(6) alignment(center) ///
	fonttbl(\f0\fnil Arial Narrow; ) 
	
	
	
	***********************************************
	*** FIGURE A1: SYMPTOMS of SECONDARY TRAUMA ***
	***********************************************

	*********
	***	Test to see if difference in STS symptoms for students from at the start
	*** of the semester and the end of the semester, following their internship
	*** experience.
	*********

ttest symptoms_numb, by(post) unequal
ttest symptoms_heart, by(post) unequal
ttest symptoms_reliving, by(post) unequal
ttest symptoms_sleep, by(post) unequal 
ttest symptoms_despondent, by(post) unequal 
ttest symptoms_upset, by(post) unequal 
ttest symptoms_antisocial, by(post) unequal 
ttest symptoms_jumpy, by(post) unequal 
ttest symptoms_lessactive, by(post) unequal 
ttest symptoms_preoccupied, by(post) unequal 
ttest symptoms_distracted, by(post) unequal 
ttest symptoms_avoidance, by(post) unequal 
ttest symptoms_dreams, by(post) unequal 
ttest symptoms_avoidsome, by(post) unequal 
ttest symptoms_annoyed, by(post) unequal 
ttest symptoms_fear, by(post) unequal 
ttest symptoms_memory, by(post) unequal

	***********
	*** RESULTS: Only results that are statistically different at alpha=.05 are  
	*** the preoccupied variable and the upset variable. The reliving variable 
	*** is stat significant at the alpha=.10. The remaining symptoms see some movement
	*** but not statistically significant. Some of this may be with stress at the end
	*** of the semester rather than STS. However, starting a new internship is pretty
	*** stressful too. 
	***********

bysort post: ci means symptoms_numb 

graph dot (mean) symptoms_numb , over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Felt Emotinally Numb", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_numb, replace)  
	
bysort post: ci means symptoms_heart 

graph dot (mean) symptoms_heart, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("My Heart Pounds When I Think of Work", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_heart, replace)  	
	
bysort post: ci means symptoms_reliving 

graph dot (mean) symptoms_reliving, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Reliving Trauma of Clients/Constituents/Customers", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_reliving, replace)  	
	
bysort post: ci means symptoms_sleep 

graph dot (mean) symptoms_sleep, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Had Trouble Sleeping", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_sleep, replace) 
	
bysort post: ci means symptoms_despondent 

graph dot (mean) symptoms_despondent, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Felt Discouraged About the Future", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_despondent, replace) 
	
bysort post: ci means symptoms_despondent 

graph dot (mean) symptoms_despondent, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Felt Discouraged About the Future", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_despondent, replace) 	
	
bysort post: ci means symptoms_upset 

graph dot (mean) symptoms_upset, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Reminders of My Work Upset Me", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_upset, replace) 	

bysort post: ci means symptoms_antisocial 

graph dot (mean) symptoms_antisocial, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Felt Antisocial", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_antisocial, replace) 	 

bysort post: ci means symptoms_jumpy 

graph dot (mean) symptoms_jumpy, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Felt Jumpy", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_jumpy, replace)  
	
bysort post: ci means symptoms_lessactive 

graph dot (mean) symptoms_lessactive, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Was Less Active Than Usual", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) /// 
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_lessactive, replace)  
		
bysort post: ci means symptoms_preoccupied 

graph dot (mean) symptoms_preoccupied, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Thought About Work When Didn't Intend To", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_preoccupied, replace) 
		
bysort post: ci means symptoms_distracted 

graph dot (mean) symptoms_distracted, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Had Trouble Concentrating", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_distracted, replace) 
		
bysort post: ci means symptoms_avoidance 

graph dot (mean) symptoms_avoidance, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Avoid People/Places/Things Reminding Me of Work", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_avoidance, replace) 
		
bysort post: ci means symptoms_dreams 

graph dot (mean) symptoms_dreams, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Had Disturbing Dreams About Work", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_dreams, replace)		
		
bysort post: ci means symptoms_avoidsome 

graph dot (mean) symptoms_avoidsome, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Wanted to Avoid Work with Some Folks", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_avoidsome, replace)		
	
bysort post: ci means symptoms_annoyed 

graph dot (mean) symptoms_annoyed, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Was Easily Annoyed", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_annoyed, replace)	

bysort post: ci means symptoms_fear 

graph dot (mean) symptoms_fear, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Expected Something Bad to Happen to Me", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_fear, replace)	

bysort post: ci means symptoms_memory 

graph dot (mean) symptoms_memory, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("I Noticed Gaps in My Memory Related to Work", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(symptoms_memory, replace)	
	
	*** Install GRC1leg to combine graphs: 
	*** package grc1leg from http://www.stata.com/users/vwiggins. 
	
grc1leg symptoms_numb.gph symptoms_heart.gph symptoms_reliving.gph ///
symptoms_sleep.gph symptoms_despondent.gph  ///
symptoms_upset.gph symptoms_antisocial.gph symptoms_jumpy.gph ///
symptoms_lessactive.gph symptoms_preoccupied.gph symptoms_distracted.gph ///
symptoms_avoidance.gph symptoms_dreams.gph symptoms_avoidsome.gph ///
symptoms_annoyed.gph symptoms_fear.gph symptoms_memory.gph, ///
cols(2) row(9)  graphregion(color(white)) plotregion(color(white)) ///
title("{bf:Figure A1: SYMPTOMS OF SECONDARY TRAUMA}") ///
saving(figurea1, replace)

graph export figurea1.svg, width(10000) replace	
	

	***********
	*** Explore exposure to secondary trauma at beginning and end of their 
	*** internship experiences.
	***********	

sum secondary_trauma_heard secondary_trauma_familiar
	
ttest secondary_trauma_exp, by(post) unequal
ttest secondary_trauma_anticipate, by(post) unequal
	
	*** 
	*** Results: Prior to the survey, respondents indicated limited awareness 
	*** of secondary trauma, averaging 0.41 on a No (0), Maybe (0.5), Yes (1) 
	*** scale. However, by the end of their internship following the surveys 
	*** and video lecture, their familiarity with the concept increased 
	*** significantly, with an average rating of 4.28 on a 1–5 scale (between 
	*** "very familiar" and "extremely familiar"). Regarding exposure to STS, 
	*** 23.3% reported experiencing it during a prior internship or their 
	*** current praxis. By the end of the semester, this figure rose to 38.4%, 
	*** a statistically significant increase (p=0.003) at the alpha=0.05 level. 
	***

***
*** Explore descriptive statistics around Respondents' familiarity
*** with STS in pre and post-survey results
***

*** How familiar are students with STS after the semester, lecture, and surveys?
*** How familiar were students with STS at the start of the semester?

tab secondary_trauma_heard
sum secondary_trauma_heard

*** install package sg113_2 from http://www.stata-journal.com/software/sj9-4
modes secondary_trauma_heard

tab secondary_trauma_familiar
sum secondary_trauma_familiar

*** RESULTS: When asked whether they had heard of STS before,
*** on a scale of No (0), Maybe (0.5), or Yes (1), The modal response for 
*** respondents regarding whether or not they had heard of STS is "No".
*** The average score was 0.41, suggest that most students were either unfamiliar
*** with or only partially familiar with STS.

*** Moreover, after their internships, when asked how familiar with they were
*** with STS on a scale of Not at all Familiar "1" to Extremely Familiar (5), the average
*** response was 4.28, indicating that were either moderately or extremely familiar
*** with STS.

*** Per reviewer recommendation, Adjust scale of familiarty question from 
*** post-test to align with 0 to 1 pre-test question, although different 
*** scale since 5 points rather than 3 in pre-test. 

gen secondary_trauma_fam_newscale = secondary_trauma_familiar
recode secondary_trauma_fam_newscale (1=0) (2=0.25) (3=0.5) (4=0.75) (5=1)
sum secondary_trauma_fam_newscale

*** Prior to or at the start of their internships, reported being somewhat
*** aware of or familiar with the term secondary trauma. The average response
*** was 0.41 on a 0 "Not Familiar", 0.5 "Maybe Familiar" 1 "Familiar" scale.
*** At the end of their internships this shifted to an average of 0.82 on a 5-point
*** Likert scale of 0 "Not Familiar" to 5 "Extremely Familiar". Although 
*** different scales, this shows a sizable increase in awareness of STS due
*** to the intervention.  

	
*** Test secondary trauma exposure by internship type. Here are the internship
*** types again: internship placement type, 1=advocacy/lobbying; 2=judicial/legal; 
*** 3=executive; 4=legislative/Congress; 5=local; 6=nonprofit; 
*** 7=campaign/political party; 8=other. Create dummy categories for these
*** different internship types

gen lobbying=0
replace lobbying=1 if intern_type==1
label variable lobbying "Internship type is lobbying internship"

gen legal=0
replace legal=1 if intern_type==2
label variable legal "Internship type is legal internship"

gen executive=0
replace executive=1 if intern_type==3
label variable executive "Internship type is Executive Branch internship"

gen legislative=0
replace legislative=1 if intern_type==4
label variable legislative "Internship type is Legislative Branch internship"

gen local=0
replace local=1 if intern_type==5
label variable local "Internship type is local government internship"

gen nonprofit=0
replace nonprofit=1 if intern_type==6
label variable nonprofit "Internship type is nonprofit sector internship"

gen campaign=0
replace campaign=1 if intern_type==7
label variable campaign "Internship type is political campaign internship"

gen other=0
replace other=1 if intern_type==8
label variable other "Internship type is other / miscellaneous internship"

sum lobbying legal executive legislative local nonprofit campaign other

***
*** Explore preliminary STS exposure by internship type
***

*** Lobbying / Advocacy Interns
ttest secondary_trauma_exp if lobbying==1, by(post) unequal 

*** Legal Interns
ttest secondary_trauma_exp if legal==1, by(post) unequal 

*** Executive Branch Interns
ttest secondary_trauma_exp if executive==1, by(post) unequal 

*** Legislative (State or Federal) Branch Interns
ttest secondary_trauma_exp if legislative==1, by(post) unequal 

*** Local Government Interns
ttest secondary_trauma_exp if local==1, by(post) unequal 

*** Nonprofit Interns
ttest secondary_trauma_exp if nonprofit==1, by(post) unequal 

*** Campaign / Political Party Interns
ttest secondary_trauma_exp if campaign==1, by(post) unequal 


*** 
*** Create a log of internship hours so extreme low or high values
*** not affecting the outcomes. Some students put in long hours
*** especially during Summer semesters. Do not want these to be influential
*** observations, so log the hours to put on similar scale. 

gen loghours=log(hours)
label variable loghours "Natural log number of internship hours"

***
*** Now, analyze STS exposure, pre vs. post with key demographic and 
*** internship characteristics / type, primary trauma and prior
*** internship controls. Use random effects.
***

******************************
*** FIGURE 1 IN MANUSCRIPT ***
******************************

xtset id time

xtreg secondary_trauma_exp age status female firstgen minority paid ///
lobbying executive legal local nonprofit campaign other ///
loghours primary_trauma intern_prior post, re vce(cluster id) 

xttest0
xttest1

*** Title: Figure 1: Political Science Interns' STS Exposure
*** Notes: Notes: Random Effects Linear Regression Model for Panel Data. 
*** Dependent Variable is Exposure to Secondary Trauma on 0 (No), 0.5 (Maybe) 
*** to 1 (Yes) scale. Robust standard errors clustered on individual IDs. Point 
*** estimates are displayed with 95% 2-tail confidence intervals. Estimates 
*** whose confidence intervals do not cross the red zero line are statistically 
*** significant at the α=0.05 level. Omitted category for internship type is 
*** legislative internships.

coefplot (, label(Coefficient) msymbol(O)) ///
	 , xline(0, lcolor(red)) legend(rows(1)) drop(_cons) scheme(s1manual) ///
	  yscale(alt noline) graphregion(margin(l=65)) ///
	  coeflabels(age="Age" status="Residency Status" ///
	female="Female" minority="Racial/Ethnic Minority" ///
	firstgen="First Generation" paid="Paid Internship" ///
	loghours="Avg. Log No. Intern Hours/Week" ///
	intern_prior="Completed Prior Internship" ///
	primary_trauma="Primary Trauma Exposure" post="Reported STS in Post Survey" ///
	lobbying="Lobbying Internship" legal="Legal Internship" ///
	executive="Executive Branch Internship" /// 
	local="Local Internship" nonprofit="Nonprofit Internship" ///
	campaign="Campaign Internship" other="Other Internship", notick labgap(-145)) ///
	nolabels ysize(8) xsize(13.25) xtitle("") legend(off) ///
	title("", size(medium)) ///
	note("", size(tiny)) ///
	order(coeflist post primary_trauma minority female intern_prior age ///
	firstgen loghours status paid ///
	lobbying legal executive local ///
	nonprofit campaign other) ///
	headings(post = "{bf:Change in STS}" primary_trauma= "{bf:Individual Characteristics}" ///
    lobbying= "{bf:Internship Type}", labgap(-150))
	
	
graph export figure1.tif, width(10000) replace


***
*** End Note 6: Random effects model vs. Fixed effects model. 
*** Test difference between random effects and fixed effects models. 
***

*** Approach this from a different angle. Consider Fixed Effects Regression

*** Install package st0113 from http://www.stata-journal.com/software/sj6-4
*** Install package package st0039 from http://www.stata-journal.com/software/sj3-2
*** Install package st0004 from http://www.stata-journal.com/software/sj1-1
*** Install package sg164_1 from http://www.stata-journal.com/software/sj8-1

xtset id time

xtreg secondary_trauma_exp post, fe vce(cluster id) 
xttest3
xtserial secondary_trauma_exp post

xtreg secondary_trauma_exp age female firstgen minority ///
loghours primary_trauma, re vce(cluster id) 
xttest0 

xtreg secondary_trauma_exp age female firstgen minority ///
loghours primary_trauma, fe vce(cluster id) 

xtreg secondary_trauma_exp age female firstgen minority ///
lobbying executive legal local nonprofit campaign other ///
loghours primary_trauma msu intern_prior post post, re 

estimates store random

xtreg secondary_trauma_exp  post, fe 

estimates store fixed 

hausman fixed random, sigmamore

*** The hausman test provides some evidence to go with a random effects model. 
*** This means there's no systematic difference between the coefficients 
*** estimated by the fixed and random effects models. Therefore, the random 
*** effects model is preferred, because it is more efficient (i.e., it gives 
*** smaller standard errors) under the null hypothesis.


******************************
*** FIGURE 2 IN MANUSCRIPT ***
******************************

*** Produce the marginal Effect

quietly xtreg secondary_trauma_exp c.age i.status i.female i.firstgen i.minority i.paid ///
i.lobbying i.executive i.legal i.local i.nonprofit i.campaign i.other ///
c.loghours i.primary_trauma i.intern_prior i.post, re vce(cluster id) 

margins, dydx(*) asobs

*** Female

margins i.female  , asobs

marginsplot,  xlabel(0(1)1) recastci(rarea) scheme(s1manual) ///
title("") xtitle("Gender") ///
ytitle("Pr(STS Exposure)") xlab(0 "Other" 1 "Female", labsize(small)) ///
saving(margins_female.gph, replace) 

margins, dydx(female)

*** Primary Trauma Exposure

margins i.primary_trauma, asobs

marginsplot, xlabel(1(1)5)  recastci(rarea) scheme(s1manual) ///
title("") xtitle("Primary Trauma Exposure") ///
ytitle("Pr(STS Exposure)") xlab(1 "None" 2 3 4 5 "Extensive", labsize(small)) ///
saving(margins_primarytrauma.gph, replace)

margins, dydx(primary_trauma)

*** Reported STS Exposure at End of Semester

margins i.post, asobs

marginsplot, xlabel(0(1)1) recastci(rarea) scheme(s1manual) ///
title("") xtitle("Survey Timing During Semester") ///
ytitle("Pr(STS Exposure)") xlab(0 "Pre" 1 "Post", labsize(small)) ///
saving(margins_post.gph, replace)

margins, dydx(post)

margins, at(age=(19(1)24)) asobs
marginsplot, xlabel (19(1)24)  recastci(rarea) scheme(s1manual) ///
title("") xtitle("Age") ///
ytitle("Pr(STS Exposure)") xlab(, labsize(small)) ///
saving(margins_age.gph, replace)

margins, dydx() at(age=(19(1)24)) 

*** Title: Figure 2: Predictive Margins of Key Variables on Secondary Trauma Exposure
*** Notes: Notes: DV represents respondents' self-reported exposure to secondary 
*** trauma on a 0 (No Exposure), 0.5 (Maybe), to 1 (STS Exposure) scale. 95% 
*** 2-tail Confidence Intervals.

grc1leg2 margins_post.gph margins_primarytrauma.gph margins_age.gph margins_female.gph , ///
cols(2) row(2) ycommon title("" , size(medsmall)) ///
graphregion(color(white)) plotregion(color(white)) ///
loff note("" ///
, size(tiny)) saving(figure2, replace)

graph export figure2.tif, width(10000)  replace	


*************************
*** TABLE A3 APPENDIX ***
*************************

eststo: xtreg secondary_trauma_exp age status female ///
firstgen minority paid loghours intern_prior primary_trauma post ///
lobbying executive legal local nonprofit campaign other /// 
, re vce(cluster id)
	
esttab using tableA3.rtf, replace varwidth(25)	///
	title("{\b Table A3: Political Science Interns' STS Exposure}") ///
	varlabels (age ///
	"Age" status "Residency Status" ///
	female "Female" minority ///
	"Racial/Ethnic Minority" firstgen ///
	"First Generation" ///
	paid "Paid Internship" ///
	loghours "Avg. Log No. Intern Hours/Week" ///
	intern_prior "Completed Prior Internship" ///
	primary_trauma "Primary Trauma Exposure" ///
	lobbying "Lobbying Internship" legal "Legal Internship" ///
	executive "Executive Branch Internship" other "Other Internship" /// 
	local "Local Internship" nonprofit "Nonprofit Internship" ///
	campaign "Campaign Internship" _cons "Constant" ///
	post "Reported STS in Post Survey") ///
	compress nomtitle nonumbers label eqlabels(none) ///
	alignment(r) stats(N chi2) se /// 
	fonttbl(\f0\fnil Arial Narrow; ) star(+ 0.10 * 0.05 ) ///
	order(post primary_trauma minority female intern_prior age ///
	firstgen loghours status paid ///
	lobbying legal executive local ///
	nonprofit campaign other) ///
	nonote ///
	addnotes("Notes: Random effects regression model for panel data. DV is self reported" ///
	"exposure to Secondary Trauma on 0 (No), Maybe (0.5), 1 (Yes) scale. Robust standard" ///
	"errors clustered on individual IDs. Standard errors in parentheses. * p<0.05," ///
	"+  p<0.10. Omitted category for internship type is legislative internships.") ///
  
eststo clear


	******
	*** Appendix Note: Evaluating Pre-test Priming Effect 
	******

	*** Check correlation between reported STS in pre-test and post-test
	*** High degree of correlation would indicate that respondents are more 
	*** likely to stick with their pre-test responses and potential priming.

preserve
	
bysort id (time): keep if _n == 1 | (_n == 2 & post == 1)	
collapse (mean) secondary_trauma_exp, by(id post)
reshape wide secondary_trauma_exp, i(id) j(post)	
corr secondary_trauma_exp0 secondary_trauma_exp1	
reshape long secondary_trauma_exp, i(id) j(post)
	
restore
	
*** Pearson correlation is 0.1493, which is a low correlation, further
*** suggesting that respondents are not beholden to their pre-test
*** response, and less risk of a pre-test sentization or priming effect.
		
*** Use Kim and Willson (2010) article to further explore pretest sensitization
*** / priming effect estimate. Start by calculating Cohen's d, 
*** where d(pre-post) = (mean_post minus mean_pre)/SD_pre

sum secondary_trauma_exp if post==0, detail
sum secondary_trauma_exp if post==1, detail

di (.3836 - .2329)/.4241193

*** Effect of pre-post for the treatment group then is .35532455
*** If we use the average 0.06 pre-test sensitization effect for long-term treated 
*** following Kim and Willson (2010) do, then we have a 
*** d_pre-post_adjusted value = d_pre-post minus d_pretest_sentization.

di .35532455 - .06

*** Thus, the adjusted pre-post effect size (accounting for priming) is .29532455

*** Now, calculate the adjusted posttest mean, using the formula provided by 
*** Kim and Willson (2010): posttest mean adjusted = (d_pre-post_adj * SDpretest) + mean_pretest

di (.29532455*.4241193) + .2329

*** The result is .35815284. This is the adjusted posttest mean STS exposure.
*** Test to see if statistically different from one another

ttesti 116 .358 .398 146 .233 .424

*** They are statistically distinguishable from one another at the alpha==.05 
*** level. Thus, even when adjusting for potential priming effects using
*** Kim and Willson's (2010) suggested pre-sentization effect adjustments, 
*** the STS exposure rates between pre- and post-tests are statistically
*** distinguishable, albeit slightly lower substantive rate of 0.125. 
*** This doesn't resolve potential priming from the intervention, although 6 
*** weeks between the pre-test, intervention, and post-tests. Still, it 
*** shows limited effect of priming from pre-test.  


*************************
*** TABLE A4 APPENDIX ***
*************************

*** DV is an ordinal variable, but much easier to interpret and make sense
*** of treatng it as a continuous variable. 
*** Robustness check between using linear model and ordinal logit model.
*** Explore comparison between random effects linear regression model
*** and ordered logit model. Since xtologit is an random effects model
*** no need to add re as an option.
*** Each coefficient represents the change in the log-odds of being in a 
*** higher category of the dependent variable for a one-unit increase in the 
*** predictor, holding other variables constant and accounting for 
*** individual-level random effects.

xtset id time

xtologit secondary_trauma_exp age status female ///
firstgen minority paid loghours intern_prior primary_trauma post ///
lobbying executive legal local nonprofit campaign other /// 
, vce(cluster id) 

estat ic

estimates store ologit_re_model

esttab ologit_re_model, eform

lincom post, or

predict pr*, pr 

*** Ordered Logit produces very similar results to random effects linear
*** model. Probability of maybe or yes response combined is .353562. 
*** Odds ratio of moving from pre to post STS exposure from start of internship
*** to end of internship is 2.832. Thus, students have a 2.832 times greater
*** odds of experiencing STS by the end of their internship. 

*** Test to see which model might be better, a linear regression model vs.
*** an ordinal logistic model, using comparison of AIC and BIC.

reg secondary_trauma_exp age status female ///
firstgen minority paid loghours intern_prior primary_trauma post ///
lobbying executive legal local nonprofit campaign other /// 
, vce(cluster id)

estat ic

ologit secondary_trauma_exp age status female ///
firstgen minority paid loghours intern_prior primary_trauma post ///
lobbying executive legal local nonprofit campaign other /// 
, vce(cluster id)

estat ic

*** AIC and BIC half of what they are under linear regression model
*** compared to ordinal logistic regressional model. Again, per (Carifio and Perla 2007),
*** "Many studies have shown that Likert Scales (as opposed to single Likert
*** response format items) produce interval data and that the F-test is very 
*** robust to violations of the interval data assumption and moderate 
*** skewing and may be used to analyze “Likert data” (even if it is ordinal)..."
*** Given this, and given that interpretation of linear regression model 
*** results is more parsimonious, opt for linear regression model.

eststo: xtologit secondary_trauma_exp age status female ///
firstgen minority paid loghours intern_prior primary_trauma post ///
lobbying executive legal local nonprofit campaign other /// 
, vce(cluster id)
	
esttab using tableA4.rtf, eform replace varwidth(25)	///
	title("{\b Table A4: Political Science Interns' STS Exposure Using RE Ordered Logit}") ///
	varlabels (age ///
	"Age" status "Residency Status" ///
	female "Female" minority ///
	"Racial/Ethnic Minority" firstgen ///
	"First Generation" ///
	paid "Paid Internship" ///
	loghours "Avg. Log No. Intern Hours/Week" ///
	intern_prior "Completed Prior Internship" ///
	primary_trauma "Primary Trauma Exposure" ///
	lobbying "Lobbying Internship" legal "Legal Internship" ///
	executive "Executive Branch Internship" other "Other Internship" /// 
	local "Local Internship" nonprofit "Nonprofit Internship" ///
	campaign "Campaign Internship" _cons "Constant" ///
	post "Reported STS in Post Survey") ///
	compress nomtitle nonumbers label eqlabels(none) ///
	alignment(r) stats(N chi2) se /// 
	fonttbl(\f0\fnil Arial Narrow; ) star(+ 0.10 * 0.05 ) ///
	order(post primary_trauma minority female intern_prior age ///
	firstgen loghours status paid ///
	lobbying legal executive local ///
	nonprofit campaign other) ///
	nonote ///
	addnotes("Notes: Random effects ordinal logit model for panel data. DV is self reported" ///
	"exposure to Secondary Trauma on 0 (No), Maybe (0.5), 1 (Yes) scale. Robust standard" ///
	"errors clustered on individual IDs. Standard errors in parentheses. * p<0.05," ///
	"+  p<0.10. Coefficients are displayed as Odds Ratios. Omitted category for" ///
	"internship type is legislative internships.")
  
eststo clear


*************************
*** TABLE A5 APPENDIX ***
*************************

*** Does experiencing STS at a higher rate lead to you establishing an action
*** plan? 

regress actionplan secondary_trauma_exp primary_trauma age status gpa minority ///
female firstgen loghours intern_prior paid lobbying legal executive /// 
local nonprofit campaign other if post==1, vce(cluster id)

*** RESULTS: Yes, there is a relationship between experiencing greater STS
*** and creating an action plan (statistically significant at alpha=.001). 
*** For every unit increase in secondary trauma exposure (0.5) from no, to maybe,
*** to yes, a 0.345 increase in likelihood of creating an action plan.

eststo: quietly regress actionplan secondary_trauma_exp primary_trauma age status gpa minority ///
female firstgen loghours intern_prior paid lobbying legal executive /// 
local nonprofit campaign other if post==1, vce(cluster id)
	
esttab using tableA5.rtf, replace varwidth(25)	///
	title("{\b Table A5: Student Create STS Action Plan}") ///
	varlabels (age ///
	"Age" status "Residency Status" ///
	female "Female" minority ///
	"Racial/Ethnic Minority" firstgen ///
	"First Generation" ///
	paid "Paid Internship" ///
	loghours "Avg. Log No. Intern Hours/Week" ///
	intern_prior "Completed Prior Internship" ///
	primary_trauma "Primary Trauma Exposure" ///
	gpa "GPA" ///
	lobbying "Lobbying Internship" legal "Legal Internship" ///
	executive "Executive Branch Internship" other "Other Internship" /// 
	local "Local Internship" nonprofit "Nonprofit Internship" ///
	campaign "Campaign Internship" _cons "Constant" ///
	secondary_trauma_exp "Reported STS in Post Survey") ///
	compress nomtitle nonumbers label eqlabels(none) ///
	alignment(r) ar2 se ///
	fonttbl(\f0\fnil Arial Narrow; ) star(+ 0.10 * 0.05) ///
	order( secondary_trauma_exp primary_trauma minority female intern_prior age ///
	firstgen loghours status paid gpa ///
	lobbying legal executive local ///
	nonprofit campaign other) ///
	nonote ///
	addnote("Linear probability model, with robust standard errors clustered on respondent ID." ///
	"Standard errors in parentheses. * p<0.05, + p<0.10. DV: Student reports creating action" ///
	"plan to mitigate Secondary Traumatic Stress on 0 (No) to 1 (Yes) scale. " ///
	"Omitted reference category for internship type is legislative internship.")
  
eststo clear


***
*** Explore changes in Physical, Psychological, Emotional, and Professional 
*** practices between pre- and post surveys.
***

*****************
*** FIGURE A2 ***
*****************

*** Physical

ttest physical_eatregularly, by(post) unequal 
ttest physical_eathealthy, by(post) unequal 
ttest physical_exercise, by(post) unequal 
ttest physical_healthcheckup, by(post) unequal 
ttest physical_sicktime, by(post) unequal 
ttest physical_sleep, by(post) unequal 
ttest physical_notechnology, by(post) unequal

*** RESULTS: Slight decreases in all of these behaviors across the board, but
*** not statistically significant. These slight declines are likely due to the
*** stress, chaos, and packed schedules at the end of the semester.

bysort post: ci means physical_eatregularly  

graph dot (mean) physical_eatregularly, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Eat Regularly", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_eatregularly, replace)  

bysort post: ci means physical_eathealthy 
 
graph dot (mean) physical_eathealthy, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Eat Healthy", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_eathealthy, replace)  

bysort post: ci means physical_exercise 
   
graph dot (mean) physical_exercise, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Exercise", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_exercise, replace)   
 
bysort post: ci means physical_healthcheckup  
 
graph dot (mean) physical_healthcheckup, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Regular Health Checkup", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_healthcheckup, replace)  

bysort post: ci means physical_sicktime  
 
graph dot (mean) physical_sicktime, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Take Time Off When Sick", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_sicktime, replace) 
 
bysort post: ci means physical_sleep 
 
graph dot (mean) physical_sleep, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Get Enough Sleep", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_sleep, replace) 
 
bysort post: ci means physical_notechnology
 
graph dot (mean) physical_notechnology, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Avoid Distracting Technology", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(physical_notechnology, replace)  
 
grc1leg physical_eatregularly.gph physical_eathealthy.gph physical_exercise.gph ///
physical_healthcheckup.gph  ///
physical_sicktime.gph physical_sleep.gph physical_notechnology.gph, ///
cols(2) row(4) pos(3) title("{bf:Figure A2: PHYSICAL SELF-CARE}") ///
graphregion(color(white)) plotregion(color(white)) ///
saving(figurea2, replace)

graph export figurea2.svg, width(5000) replace

*****************
*** FIGURE A3 ***
*****************

*** Psychological

ttest psychological_therapy, by(post) unequal  
ttest psychological_stress, by(post) unequal  
ttest psychological_dreams, by(post) unequal  
ttest psychological_cultural, by(post) unequal  
ttest psychological_sayno, by(post) unequal  
ttest psychological_outdoors, by(post) unequal 

*** RESULTS: Again, no real statistical difference between pre- and post EXCEPT
*** for the psychological stress question: How often do you (5=very often, 1=never)
*** Take a step to decrease stress in your life? Students reported an increase
*** in trying to take active steps to decrease stress in their lives.

bysort post: ci means psychological_therapy  

graph dot (mean) psychological_therapy, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Visit Therapist", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_therapy, replace)  

bysort post: ci means psychological_stress 
 
graph dot (mean) psychological_stress, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Decrease Stress", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_stress, replace)  

bysort post: ci means psychological_dreams 
   
graph dot (mean) psychological_dreams, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Notice Inner Self", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_dreams, replace)   
 
bysort post: ci means psychological_cultural  
 
graph dot (mean) psychological_cultural, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Attend Cultural Events", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_cultural, replace)  

bysort post: ci means psychological_sayno  
 
graph dot (mean) psychological_sayno, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Say No to Extra Responsibilities", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_sayno, replace) 
 
bysort post: ci means psychological_outdoors 
 
graph dot (mean) psychological_outdoors, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Spend Time Outdoors", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(psychological_outdoors, replace) 
 
grc1leg psychological_therapy.gph psychological_stress.gph psychological_dreams.gph ///
psychological_cultural.gph  ///
psychological_sayno.gph psychological_outdoors.gph, ///
cols(2) row(2) pos(3) title("{bf:Figure A3: PSYCHOLOGICAL SELF-CARE}") ///
graphregion(color(white)) plotregion(color(white)) ///
saving(figurea3, replace)

graph export figurea3.svg, width(5000) replace

*****************
*** FIGURE A4 ***
*****************

*** Emotional

ttest emotional_connect, by(post) unequal 
ttest emotional_outrage, by(post) unequal
ttest emotional_selftalk, by(post) unequal 
ttest emotional_comforts, by(post) unequal 
ttest emotional_cry, by(post) unequal 
ttest emotional_laugh, by(post) unequal 
ttest emotional_outrage, by(post) unequal 
ttest emotional_emotions, by(post) unequal

*** RESULTS: No real changes in any of these emotional metrics pre and post. 

bysort post: ci means emotional_connect  

graph dot (mean) emotional_connect, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Contact Important People in Life", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_connect, replace)  

bysort post: ci means emotional_outrage  

graph dot (mean) emotional_outrage, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Express Outrage Constructively", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_outrage, replace)   
 
bysort post: ci means emotional_selftalk 
 
graph dot (mean) emotional_selftalk, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Supportive Self-Talk", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_selftalk, replace)  

bysort post: ci means emotional_comforts 
   
graph dot (mean) emotional_comforts, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Identify & Seek Comforting Activities", size(medlarge) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_comforts, replace)   
 
bysort post: ci means emotional_cry  
 
graph dot (mean) emotional_cry, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Allow Yourself to Cry", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_cry, replace)  

bysort post: ci means emotional_laugh  
 
graph dot (mean) emotional_laugh, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Find Things to Make Yourself Laugh", size(medlarge) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(emotional_laugh, replace) 
 
bysort post: ci means emotional_emotions
 
graph dot (mean) emotional_emotions, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Process Emotions Opening Up to Others", size(medlarge) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often")  ///
 saving(emotional_emotions, replace)  
 
grc1leg emotional_connect.gph emotional_selftalk.gph emotional_comforts.gph ///
emotional_cry.gph  ///
emotional_laugh.gph emotional_outrage.gph emotional_emotions.gph, ///
cols(2) row(3) pos(3) title("{bf:Figure A4: EMOTIONAL SELF-CARE}") ///
graphregion(color(white)) plotregion(color(white)) ///
saving(figurea4, replace)

graph export figurea4.svg, width(5000) replace

*****************
*** FIGURE A5 ***
*****************

*** Professional

ttest professional_chat, by(post) unequal 
ttest professional_time, by(post) unequal 
ttest professional_growth, by(post) unequal 
ttest professional_support, by(post) unequal 
ttest professional_peer, by(post) unequal

*** RESULTS: The only two categories that appear to be statistically distinguishable
*** between pre and post tests are the Professional chat (alpha=.10), and 
*** Professional growth. The chat question is "How often do you take time to chat
*** with coworkers (5=very often, 1=never)? There is an increase in this response
*** from 3.82 to 4.02 (alpha = .078). And the growth question is "How often do you
*** Identify with projects or tasks that are exciting, growth promoting, and rewarding
*** for you (5=very often, 1=never)? There is a reported increase from 3.81 to
*** 4.03 (alpha=.0461). Statistically significant at alpha=.05 level.


bysort post: ci means professional_chat  

graph dot (mean) professional_chat, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Chat with Coworkers", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(professional_chat, replace)  

bysort post: ci means professional_time 
 
graph dot (mean) professional_time, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Make Time to Complete Tasks", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(professional_time, replace)  

bysort post: ci means professional_growth 
   
graph dot (mean) professional_growth, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Identify Exciting/Rewarding Projects", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(professional_growth, replace)   
 
bysort post: ci means professional_support  
 
graph dot (mean) professional_support, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Get Support from Colleagues", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(professional_support, replace)  

bysort post: ci means professional_peer
 
graph dot (mean) professional_peer, over(post, relabel(1 "{bf:Pre}" 2 "{bf:Post}")) ///
 asyvar missing nofill m(1, m(O) msize(large) mlcolor(black) mfcolor(black) mlw(medium)) ///
 m(2, m(Oh) msize(large) mlcolor(black) mfcolor(blue) mlw(medium)) legend(off) exclude0 ///
 title("Have Peer Support Group", size(large) margin(zero)) ///
 ytitle("") ysc(r(1 5)) ymtick(1(1)5) ///
 graphregion(margin(r+5)) plotregion(margin())  ///
 graphregion(color(white)) plotregion(color(white)) ///
 ylabel(1 "Never" 2 3 4 5 "Very Often") ///
 saving(professional_peer, replace) 
 
grc1leg professional_chat.gph professional_time.gph professional_growth.gph ///
professional_support.gph  ///
professional_peer.gph, ///
cols(2) row(3) pos(3) title("{bf:Figure A5: WORKPLACE SELF-CARE}") ///
graphregion(color(white)) plotregion(color(white)) ///
saving(figurea5, replace)

graph export figurea5.svg, width(5000) replace

	*********************
	*** Save the Data ***
	*********************

save secondary_trauma_ps_replication.dta, replace


	********
	*** Close Log
	********

log close
